HIGHER-DIMENSIONAL FROBENIUS GAPS
|
|
- Ethelbert Atkinson
- 5 years ago
- Views:
Transcription
1 HIGHER-DIMENSIONAL FROBENIUS GAPS A thesis presented to the faculty of San Francisco State University In partial fulfilment of The requirements for The degree Master of Arts In Mathematics by Jessica Delgado San Francisco, California July 2013
2 Copyright by Jessica Delgado 2013
3 CERTIFICATION OF APPROVAL I certify that I have read HIGHER-DIMENSIONAL FROBENIUS GAPS by Jessica Delgado and that in my opinion this work meets the criteria for approving a thesis submitted in partial fulfillment of the requirements for the degree: Master of Arts in Mathematics at San Francisco State University. Matthias Beck Professor of Mathematics Joseph Gubeladze Professor of Mathematics Serkan Hosten Professor of Mathematics
4 HIGHER-DIMENSIONAL FROBENIUS GAPS Jessica Delgado San Francisco State University 2013 This research expands on the well studied Frobenius problem and examines a related problem. This thesis mainly focuses on very ample polytopes of dimension three and their gaps, lattice points in the homogenization of the polytope that cannot be written as integer combinations of lattice points in the poytope. The main result is a theorem which states a universal upper bound of the gaps of very ample polytopes in dimension three does not exist. We built a program to compute the gaps of any very ample polytope. The computations of explicit examples are used to prove the main result on the nonexistence of the upper bound as well as a conjecture on the behavior of the gaps. I certify that the Abstract is a correct representation of the content of this thesis. Chair, Thesis Committee Date
5 ACKNOWLEDGMENTS I d like to thank Dr. Matthias Beck and Dr. Joseph Gubeladze, your challenges and endless patience have made me into the mathematician I am today. I would also like to thank Winfried Bruns for sharing his library of very ample polytopes, they contributed greatly to this research. v
6 TABLE OF CONTENTS 1 Introduction Frobenius Problem Monoids and Numerical Semigroups Polytopes Cones Dimension two Condition for a finite number of gaps Where the gaps are located Very Ample Polytopes What is a very ample polytope? Non-Normal Very Ample Polytope of Dimension d With h Gaps Polytopal Affine Semigroups With Gaps Deep Inside How to find gaps of very ample polytopes Bruns Polytopes Very ample polytopes in dimension Is there a universal bound for the gaps? Appendix A: JAVA code to generate gaps for a given a Appendix B: Python code that computes gaps of very ample polytopes vi
7 Appendix C: R formatting program Bibliography vii
8 LIST OF TABLES 3.1 Gaps of Bruns polytopes Gap heights of various BG polytopes Gaps of BG polytope when I 1 = I 2 = I 3 = {0, 1 and I 4 = {k, k viii
9 LIST OF FIGURES 1.1 The shaded area is H + α The convex hull of {a, b, c, d, e, f C(P) The first parallelogram with verticies (0, 0), (1, 1), (1, 2), (0, 1) The darker dots are the gaps The dots are the gaps The gap behavior of a very ample polytope P I 1 = {0, 1 I 2 = {0, 1 I 3 = {0, 1 I 4 = {4, ix
10 Chapter 1 Introduction 1.1 Frobenius Problem In the late nineteenth century Ferdinand Georg Frobenius gave the following problem to his students: given relatively prime positive integers a 1, a 2,..., a n find the largest natural number such that is not representable as a non-negative integer linear combination of a 1, a 2,..., a n [11]. Definition 1.1. Given A = {a 1, a 2,..., a n Z +, an integer m is representable if m = x 1 a 1 + x 2 a x n a n where x i Z 0 and non-representable otherwise. is For instance if A = {5, 7, 9 then the set of non-representable positive integers {1, 2, 3, 4, 6, 8, 11, 13. The largest non-representable number in this case is 13. This 1
11 2 largest natural number in the general case became known as the Frobenius number, denoted as F (a 1, a 2,..., a n ). This problem has also been referred to as the coin exchange problem. Suppose we abandoned our current coin system and only used 3 cent and 7 cent pieces. Although this seems foolish, this new system poses an interesting problem: there are amounts we cannot make change for. What is the largest amount that cannot be changed? Our set of unchangeable amounts are {1, 2, 4, 5, 8, 11. Therefore if we changed our coin system to only 3 and 7 cent pieces the largest amount we could not make change for is 11 cents. In general the problem goes as follows: a 1, a 2,..., a n are relatively prime coin denominations, find the largest amount that you cannot make change for. More recently in the 1980s, there was a twist put on the problem by coming up with the McNugget number. McDonalds originally sold chicken nuggets in packages of 6, 9, and 20, and you could only order representable amounts of McNuggets. The largest number of McNuggets that you cannot order from McDonalds is precisely F (6, 9, 20) = 43 [14]. Over 100 years mathematicians from different areas have been researching this problem. The main area of interest is trying to produce formulas or algorithms for F (a 1, a 2,..., a n ). So far there has been success for producing a formula for n = 2; beyond this, however, there has been computational algorithms that run in polynomial time and formulas for bounds [11]. The following is the closed formula
12 3 for n = 2 [11]. Theorem 1.1. Given a 1, a 2 N such that gcd(a 1, a 2 ) = 1 then F (a 1, a 2 ) = a 2 a 1 a 1 a 2. Proof. Since gcd(a 1, a 2 ) = 1 then any integer p can be written as p = xa 1 + ya 2 with x, y Z. It can be seen that p can be expressed in many different ways however the expression becomes unique when 0 x < a 2. In this case p is representable if y 0 and not if y < 0. Therefore the largest non-representable integer will occur when x = a 2 1 and y = 1. Thus F (a 1, a 2 ) = (a 2 1)a 1 + ( 1)a 2 = a 2 a 1 a 1 a 2. It is also known how many of the integers will be representable. Sylvester came up with a theorem for the amount of representable integers when n = 2 [1]. Theorem 1.2. (Sylvester s theorem) Let a 1 and a 2 be relatively prime positive integers. Exactly half of the integers between 1 and (a 1 1)(a 2 1) are representable. To prove this theorem we will use the following lemma [1]. Lemma 1.3. If a 1 and a 2 are relatively prime positive integers and n [1, a 1 a 2 1] is not a multiple of a 1 or a 2 then exactly one of the two integers n and a 1 a 2 n is representable in terms of a 1 and a 2.
13 4 Proof of Theorem 1.2. From Lemma 1.3 we know that for n between 1 and a 1 a 2 1 and not divisible by a 1 or a 2 exactly one of n and a 1 a 2 n is representable. There are a 2 1 numbers divisible by a 1 and a 1 1 numbers divisible by a 2. Therefore there are (a 1 a 2 1) (a 1 1) (a 2 1) = a 1 a 2 a 1 a = (a 1 1)(a 2 1) integers between 1 and a 1 a 2 1 that are not divisible by a 1 and a 2. Since exactly one of n or a 1 a 2 n is representable, the number of non-representable integers is 1 (a 2 1 1)(a 2 1). In the case of n = 3 there has been much less success coming up with a closedform expression, instead there are a couple of well-known algorithms such as Davison s algorithm [6] and Rødseth s algorithm [13]. There has in fact been proof that one cannot come up with an explicit formula for n > 2 by F. Curtis [5]. Theorem 1.4. Let A = {(a 1, a 2, a 3 ) N 3 a 1 < a 2 < a 3, a 1 and a 2 are prime, and a 1, a 2 a 3 then there is no nonzero polynomial H C[X 1, X 2, X 3, Y ] such that H(a 1, a 2, a 3, F (a 1, a 2, a 3 )) = 0 for all a 1, a 2, a 3 A. The following corollary shows that F (a 1, a 2, a 3 ) cannot be determined by any set of closed formulas that could be reduced to a finite set of polynomials when restricted to A.
14 5 Corollary 1.5. There is no finite set of polynomials {h 1,..., h n such that for each choice of a 1, a 2, a 3 there is some i such that h i (a 1, a 2, a 3 ) = F (a 1, a 2, a 3 ). Proof. The polynomial H = n i=1 (h i(x 1, X 2, X 3 ) Y ) would vanish on A. The Frobenius problem is fascinating in its own right, however this problem is the inspiration and not the focus of this paper. What do the non-representable integers look like if we lift this problem into a higher dimension? The answer to that question as well as the behavior of the higher dimensional non-representable integers will be the focus of this thesis. The rest of the introduction will focus on the background knowledge required for the main results. 1.2 Monoids and Numerical Semigroups Definition 1.2. A numerical semigroup S is a subset of N that is closed under addition, contains 0, and whose complement in N is finite. If A = {a 1, a 2,..., a n contains positive integers with gcd(a 1, a 2,..., a n ) = 1 then the set generated by A is {m 1 a 1 + m 2 a m n a n m 1,..., m n N, a numerical semigroup. Every numerical semigroup is of this form [12]. A trivial example of a numerical semigroup is N {0. A non-trival example is the set {0, 3, 5, 6, 7, 8, 9, 10,.... In this example the integers 1, 2 and 4 are what I will refer to as gaps. A non-example is the set {0, 1, 3, 4, 5, 6,... because 1 is in the set but = 2 is not.
15 6 Definition 1.3. An integer q is a gap if q N \ S. We can now see how the Frobenius problem relates to numerical semigroups, the set of representable integers is in fact a numerical semigroup. If we go back to our first example where A = {5, 7, 9 then our numerical semigroup S generated by A is {5, 7, 9, 10, 12, 14, 15, 16, 17, 18,... and the set of gaps is precisely the set of non-representables {1, 2, 3, 4, 6, 8, 11, 13. Note that if x is a gap then so are all the non-negative integers that divide it. We can also see that the largest gap equals to the Frobenius number of the set A. Numerical semigroups live in the world of commutative monoids. Definition 1.4. A commutative monoid is a set M paired with a binary operations + that satisfies the following axioms: 1. For all a, b M a + b is also in M. 2. For all a, b M a + b = b + a. 3. For all a, b, c M (a + b) + c = a + (b + c). 4. There exists an element e in M such that for all elements a M e + a = a + e = a. Examples of monoids are {0, (Z 0, +), (Q, +) and (Z 0, ) Notice that elements of a monoid do not necessarily have inverses. A monoid where every element has an inverse is a group.
16 7 1.3 Polytopes Definition 1.5. Let V be a vector space. A V is an affine subspace if A = U +v = {u+v u U for some linear subspace U V and v V. Affine subspaces can be viewed as a linear subspace with a shift. Definition 1.6. Let V 1 and V 2 be vector spaces, A 1 V 1 and A 2 V 2 affine spaces. A map f : A 1 A 2 is affine if for some linear map φ : V 1 V 2 and v V 2 the following diagram commutes: V 1 φ + v V 2 f A 1 A 2 Definition 1.7. Let V be a vector space and X V finite. The affine hull aff(x) is the smallest affine subspace containing X. This is { n λi aff(x) = aff(x 1, x 2,..., x n ) = λ i x i R, λi = 1. Definition 1.8. An affine half-space, denoted as either H α + or Hα is the set H α + = {x V : α(x) b or Hα = {x V : α(x) b where b R and α : V R is a linear map. The set that is defined by {x V : α(x) = b is a hyperplane. An example of a hyperplane H can be seen in Figure 1.1. In this case the hyperplane is the line H. i=1
17 8 Figure 1.1: The shaded area is H + α. Definition 1.9. A polyhedron is a finite intersection of half-spaces, n i=1 H+ α i where α i : V R are affine maps. Notice that a polyhedron does not necessarily have to be bounded. Definition A bounded polyhedron is a polytope. Definition Let V be a vector space and X = {x 1, x 2,..., x n V. The convex hull of X is { n xi conv(x) = λ i x i X, λ i 0, λi = 1. This is the smallest convex set containing {x 1, x 2,..., x n. i=1 A polytope can be described in two different ways: there is the H-description which is the intersection of half-spaces or the V -description which is the convex hull of a set of vertices [3]. For example, a simple two-dimensional triangle can be
18 9 Figure 1.2: The convex hull of {a, b, c, d, e, f. described the following ways: conv((0, 0), (0, 1), (1, 0)) or x 0 (x, y) R 2 : y 0. y + x 1 Definition A face of a polytope is defined as the intersection of the polytope with an affine hyperplane such that H + α or H α contains the polytope. Definition Given a polytope P, a hyperplane that has the property that H + α or H α contains P is a support hyperplane.
19 10 is d: Faces of a polytope P include and P itself. Suppose that the dimension of P 0-dimesional faces: vertices 1-dimensional faces: edges. d 2 dimensional faces: ridges d 1 dimensional faces: facets d-dimensional face: polytope According to [3], any face of a polytope P is a polytope itself whose faces come from the lower-dimensional faces of P. 1.4 Cones Definition Let V be a vector space and {x 1, x 2,..., x n be a collection of vectors in V. The cone spanned by {x 1, x 2,..., x n, denoted by C(x 1, x 2,..., x n ), is the set { n i=1 λ ix i λ i 0 V. Notation. R + X = { n i=1 α ix i α i R +, x i X. Theorem 1.6. C is a cone if R + X = C for some finite non-empty set X V [3].
20 11 Definition We say that C is a simplicial cone if and only if C = R + {x 1, x 2,..., x n such that {x 1, x 2,..., x n is linearly independent. Definition Let C be a cone. We say that C is pointed if C contains no non-zero linear subspaces. Unless otherwise stated all cones in this paper will be pointed. A cone is also used in conjunction with polytopes. A polytope can generate a cone in the following way: C(P ) = R + (P, 1). Here (P, 1) = {(x, 1) x P and thus C(P ) refers to the cone of P by raising the polytope into the next dimension at height one and taking all the positive linear combinations. Figure 1.4 shows a cone generated by a polytope. Notation. For A = {a 1, a 2,..., a n, let Z + (A, 1) = {λ 1 (a 1, 1) + + λ n (a n, 1) λ i Z +.
21 Figure 1.3: C(P). 12
22 Chapter 2 Dimension two In this chapter we will examine the gaps of the special case of dimension two. There are a couple differences already between this and the one-dimensional Frobenius problem: the first, our gaps are now lattice points or points with integral coordinates and second, we will be using linear combinations of vectors instead of integers. This next dimension offers a new challenge: how to ensure a finite amount of gaps? This question and the question of how to locate the gaps will be answered in this chapter. 2.1 Condition for a finite number of gaps Definition 2.1. Let A = {a 1, a 2,..., a n Z +. A lattice point in Z 2 is a gap if it is an element of R + (A, 1) \ Z + (A, 1). Theorem 2.1. The number of gaps in R + (A, 1) \ Z + (A, 1) are finite if and only 13
23 14 if a 1, a 2,..., a n 1, a n Z 0 such that a 1 < a 2 < < a n 1 < a n and a 2 a 1 = a n a n 1 = 1 In order to prove Theorem 2.1 we want to show that at some height the gaps disappear and furthermore every height above will have no gaps as well. We first need the following lemma which will allow us to shift our set to the origin. Lemma 2.2. For A = {a 1, a 2,..., a n Z and B = {0, a 2 a 1, a 3 a 1,..., a n a 1 a i A, R + (A, 1) = R + (B, 1) R + (A, 1) Z 2 = R+ (B, 1) Z 2 Z + (A, 1) = Z + (B, 1). Here X = Y means X is isomorphic to Y as semigroups. Proof. Using {(a 1, 1), (a 2, 1) as a basis of R 2 as well as {(0, 1), (a 2 a 1, 1), we define a linear map φ : R 2 R 2 through φ(λ 1 (a 1, 1) + λ 2 (a 2, 1)) = λ 1 (0, 1) + λ 2 (a 2 a 1, 1). In order to prove that φ is injective we shall show that nullφ = {0. 0 = 0(0, 1) + 0(a 2 a 1, 1)
24 15 φ 1 (0(0, 1) + 0(a 2 a 1, 1)) = 0(a 1, 1) + 0(a 2, 1) = 0. In order to prove that φ is surjective let z R 2. Then φ 1 (z ) = φ 1 (λ 1 (0, 1) + λ 2 (a 2 a 1, 1)) = λ 1 (a 1, 1) + λ 2 (a 2, 1), which is some element of the domain R 2, therefore our range is all of R 2. We now know that φ is a linear, injective and surjective map. In order to prove R + (A, 1) = R + (B, 1), we shall use φ R+ (A,1) : R + (A, 1) R 2. We already know that φ R+ (A,1) is linear and since R + (A, 1) is a subset of R 2 it is also injective. It remains to show that the range of φ R+ (A,1) = R + (B, 1). Because of linearity, φ R+ (A,1) j=1 ( n ) λ j (a j, 1) = n λ j φ R+ (A,1)(a j, 1). j=1 We claim that φ R+ (A,1)(a j, 1) = (a j a 1, 1) giving us our desired range. Solving (a j, 1) = λ(a 1, 1) + π(a 2, 1) for λ and π gives us the system of equations λa 1 + πa 2 = a j λ + π = 1.
25 16 Setting π = 1 λ we obtain: λ = a j a 2 a 1 a 2 π = 1 a j a 2 a 1 a 2 φ(a j, 1) = a ( j a 2 (0, 1) + 1 a ) j a 2 (a 2 a 1, 1) a 1 a 2 a 1 a ( 2 = 0, a ) ( ( ) j a 2 aj a 2 + (a 2 a 1 ) + (a 1 a 2 ), 1 a ) j a 2 a 1 a 2 a 1 a 2 a 1 a ( 2 = 0, a ) ( j a 2 + a 2 a 1 + a j a 2, 1 a ) j a 2 a 1 a 2 a 1 a ( 2 = 0, a ) ( j a 2 + a j a 1, 1 a ) j a 2 a 1 a 2 a 1 a 2 = (a j a 1, 1) Therefore φ R+ (A,1)(R + (A, 1)) = { n j=1 λ j(a j a 1, 1) = R + (B, 1). We now have φ R+ (A,1) : R + (A, 1) R + (B, 1) as a linear bijective map. This shows that R + (A, 1) = R + (B, 1). In order to prove R + (A, 1) Z 2 = R+ (B, 1) Z 2, we shall use φ R+ (A,1) Z 2 : R + (A, 1) Z 2 R + (B, 1). We already know that φ R+ (A,1) Z 2 is linear and since R + (A, 1) Z 2 is a subset of R + (A, 1) it is also injective. It remains to show that the range of φ R+ (A,1) Z 2 = R +(B, 1) Z 2. Using linearity and our previous result, { n { n { n φ R+ (A,1) Z 2 λ j (a j, 1) = λ j φ R+ (A,1) Z 2(a j, 1) = λ j (a j a 1, 1). j=1 j=1 j=1
26 17 An element x R + (B, 1) Z 2 has the form (λ 2 + n j=3 λ j(a j a 1 ), n j=1 λ j) such that n j=1 λ j Z and λ 2 + n j=3 λ j(a j a 1 ) Z. We know that φ R+ (A,1) : R + (A, 1) R + (B, 1) Z 2 is surjective since R + (B, 1) Z 2 R + (B, 1). In order to { n show that j=1 λ j(a j a 1, 1) = R + (B, 1) Z 2 we must show that n j=1 λ j Z and a j a 1 Z. Since both a j and a 1 are define to be integers a j a 1 Z. The n j=1 λ j Z is not affected by φ R+ (A,1) Z2 and therefore will remain in Z. Now we show Z + (A, 1) = Z + (B, 1). There exists a bijection Φ : A B such that a i a i a 1. Therefore Z + (A, 1) = Z + (B, 1). Now that we know we can shift our set to the origin we prove there are no gaps in C((a 1, 1), (a 2, 1)) and C((a n 1, 1), (a n, 1)) with the next lemma. Lemma 2.3. For A = {a 1, a 2, a 1, a 2 Z + such that a 1 < a 2 and a 2 a 1 = 1, no gaps exist in R + (A, 1) \ M(A). Proof. Using Lemma 2.2, subtracting a 1 from each entry of A gives A = {0, 1 and (A, 1) = {(0, 1), (1, 1). Our R + (A, 1) Z 2 in this case is {λ 1 (0, 1) + λ 2 (1, 1) λ i 0 Z 2. The integer span of these two vectors is Z + (A, 1) = {v 1 (1, 1) + v 2 (0, 1) v i Z 0. We can tile R + (A, 1) Z 2 with parallelograms such that the first parallelogram has vertices at (0, 0), (1, 1), (1, 2), (0, 1), as shown in Figure 2.1. Every other parallelogram can be represented with the vertex set {(p, p), (p, p + 1), (p + 1, p +
27 18 1), (p+1, p+2) where p = 0, 1, 2,.... If every element of R + (A, 1) Z 2 is contained Figure 2.1: The first parallelogram with verticies (0, 0), (1, 1), (1, 2), (0, 1). in the vertex set of a translated copy of our first parallelogram then every element of R + (A, 1) Z 2 is covered by an element from Z + (A, 1). Therefore it is enough to show that there does not exist any integer lattice points in the interior of our first parallelogram. Suppose there was an integer lattice point (p, q) in the interior of our upper triangle. Then (p, q) would also be an integer vector and the vector (1, 2) (p, q) would also be an integer vector however if this were true that would imply that (1, 2) (p, q) leads to an integer lattice point in the lower triangle which cannot be true. We will use these two lemmas to show that after some height every lattice point
28 19 within R + (A, 1) can be written as a linear combination of points from our gap-free cones. This will ensure that from some height and above there will be no more gaps. Proof of Theorem 2.1. Using Lemma 2.2 and the set {0, 1, a, a + 1, the minimum amount of elements that meets the criteria of Theorem 2.1, we shall prove the theorem. If we lift this set into dimension two we get {(0, 1), (1, 1), (a, 1), (a + 1, 1). By Lemma 2.3 we know R + {(0, 1), (1, 1) Z 2 \ Z + {(0, 1), (1, 1) = and R + {(a, 1), (a + 1, 1) Z 2 \ Z + {(a, 1), (a + 1, 1) =. At an arbitrary height n, which is simply the second coordinate of a lattice point, the lattice points in Z + (A, 1) are {(0, n),..., (n, n) {(a, n),..., (a + n, n) {(na, n),..., (na + n, n). If we let n = a this set becomes {(0, a),..., (a, a) {(a, a),..., (2a, a) {(a 2, a),..., (a 2 + a, a). For height a these are all the lattice points between (0, a) and (a 2, a) which are
29 20 the two extremal points of the cone. At this point the cones C((0, 1), (1, 1)) and C((a, 1), (a + 1, 1)) intersect and by Lemma 2.3 which implies there are no gaps within these cones everything beyond this height will be covered. 2.2 Where the gaps are located In the previous section we proved a theorem that guaranteed a finite amount of gaps as long as A met a certain criteria. In this section we will use Theorem 2.1 and some programming to show where these gaps are located. The gaps we are locating are generated by the set A = {0, 1, a, a + 1 this way the programming can be done with respect to one parameter, a. The first step in the process was to write a JAVA program which is located in Apendix A. The idea behind the program is to find all lattice points in R + (A, 1) Z 2 up to a bound and then test each point to see if it is in Z + (A, 1). The program then returns all the points of R + (A, 1) Z 2 \ Z + (A, 1), or our gaps. This data was then reformatted and input into R to generate a graphic. Figures 2.2 and 2.3 are pictures of the gaps for a = 5 and 20 respectively.
30 21 a=5 y x Figure 2.2: The darker dots are the gaps.
31 22 a=20 y x Figure 2.3: The dots are the gaps.
32 Chapter 3 Very Ample Polytopes In the last chapter we showed how to guarantee a finite amount of gaps in dimension two and where these gaps were located. In this chapter, with the help of algebraic geometry, we will show the condition for a finite amount of gaps for any dimension. We will also show how to find these gaps. 3.1 What is a very ample polytope? In this section we will define what a very ample polytope is and why we care about them. Before we do this we will need some tools from algebraic geometry. We will be working in an affine space which is the set of all n-tuples of numbers in a field K and we will denote it as A n. For our field we will be using C because the complex numbers are algebraically closed, every polynomial has a root in C. The 23
33 24 next definitions will give some needed information about C-algebras. Definition 3.1. A ring A is a C-algebra if it is a ring and a C-vector space in a compatible way. Note that A contains an isomorphic copy of C. Definition 3.2. A C-algebra A is finitely generated if there exist finitely many elements a 1, a 2,..., a n A, called a generating set, such that A is the smallest C-subalgebra of A containing {a 1, a 2,..., a n. A C-algebra A is finitely generated if and only if A is the smallest set containing C and {a 1, a 2,..., a n and every element a A can be represented as a = n i=1 c i a d i 1 1 a d in n, c i C, d ij 0. Note that generating sets in general are not unique [7]. Notation. The C-algebra generated by a 1, a 2,..., a n is denoted by C[a 1, a 2,..., a n ]. Definition 3.3. Let x 1, x 2,..., x n be variables, which we regard as coordinate functions on A n and C[x 1, x 2,..., x n ] be the polynomial ring with coefficients in C. We may evaluate a polynomial f C[x 1, x 2,..., x n ] at a point a A n to get a number f(a) C. An example of this is if f = x 2 1 x 3 2 then f(0, 0) = 0. As one could notice (0, 0) is not the only point in A 2 that would make f(x 1, x 2 ) = 0, in fact every time x 2 = 3 x 2 1 this would be true. The points in which a given f = 0 are of importance to us.
34 25 Definition 3.4. An affine variety is the set of common zeros of a collection of polynomials. Given F C[x 1, x 2,..., x n ], let V (F ) = {a A n f(a) = 0 for all f F. Some trivial examples of affine varieties are V (1) = and V (0) = A n. Also any linear or affine subspace L of A n is a variety. We know that L has a defining equation Ax = b where A is a matrix and b is a vector so L = V (Ax b) is defined by the linear polynomials which form the rows of Ax b. Notation. Starting with a finitely generated C-algebra A, the variety that is generated by A is the affine spectrum of A, denoted as Spec(A). Definition 3.5. F C[x 1,..., x d ] is a homogeneous polynomial if F = i c i x a i x a id d such that for all i, a i a id = k, for some fixed integer k. Definition 3.6. The projective space P d is the set of one-dimensional linear subspaces of A d+1, which is equivalent to C d+1. A point in P n has homogeneous coordinates (a 0, a 1,..., a n ), where a i C are not all zero, and another set of coordinates (b 0, b 1,..., b n ) give the same point in P n if and only if there is a complex number λ 0 such that (b 0, b 1,..., b n ) = λ(a 0, a 1,..., a n ). Also if F (x 0, x 1,..., x n ) is a homogeneous of degree d and (b 0, b 1,..., b n ) = λ(a 0, a 1,..., a n ) are two sets of
35 26 homogeneous coordinates for some point p P n, then F (b 0, b 1,..., b n ) = λ d F (a 0, a 1,..., a n ). By using d + 1 affine charts from the (d + 1)-dimensional affine space, a d- dimensional projective space can be covered. In order to cover a d-dimensional projective space with affine charts, start with a homogeneous polynomial F C[x 1,..., x d ] and let V (F ) P d = {(x 0,..., x d ) F (x 0,..., x d ) = 0. Note that V (F 1,..., F n ) = n i V (F i ). Denote A i = P n \ V (x i ) therefore A i = {a 0,..., a i 1, 1, a i+1,... a d = A d. Since A i = A d then A i = Spec(C[ x 0 x i,... x i 1 x i, x i+1 x i,..., x d x i ]) = A d. Therefore P d = d i=0 Spec(C[ x 0 x i,..., x d x i ]) where Spec(C[ x 0 x i,..., x d x i ]) are the affine charts. This construction will be used in the proof of Theorem 3.3. Definition 3.7. J C[x 1,..., x d ] is a homogeneous ideal if for all F J, F = F k 1 +F k F kn where F k i are homogeneous parts of degrees k 1, k 2,..., k n. Note that J is a homogeneous ideal if and only if it is generated by homogeneous polynomials.
36 27 Theorem 3.1 (Projective Nullstellensatz). Let J C[x 1,..., x d ] where J = (F 1,..., F n ) is a homogeneous ideal. Then the following are true [4]: 1. V (J) = V ( J). 2. V (J) = x a 1 0,..., x a d d J for some a 1,..., a d N. Note: (x 1,... x d ) C[x 1,..., x d ] is called the irrelevant maximal ideal. Definition 3.8. A graded C algebra is of the form A = C A 1 A 2 where each A i is a C-vector space and A i A j A i+j. We will now begin to discuss varieties that live inside to P d, denoted as Proj(A) when A is a graded C-algebra. Definition 3.9. A finitely generated graded C-algebra, A = C A 1 A 2... is generated in degree one with dim A 1 = d, {a 1,..., a d A 1 being a C basis. Suppose A is a finitely generated C-algebra then Proj(A) P d 1 is the projective variety Proj(A) = V (J) where J = ker(c[x 1,..., x d ] A) such that x i a i. If A = C A 1 A 2... is a graded finitely generated C algebra generated in degree one then Spec(A) is the unions of the one-dimensional linear subspaces in A d, d = dim A 1, representing the points of Proj(A) P d 1.
37 28 Just as P d is covered by affine charts Proj(A) is as well. Let A be as above, then the description of the affine charts of the projective varieties in terms of ideals are A (ai ) = C A 1 A 2 A 3... A a 1 a 2 1 a 3 ai 1 where A (ai ) is the localization at a i and A i a i 1 = { a a i a A i. Let A (ai ) be a finitely generated C algebra A (ai ) = C[ a 1 a i,..., a i 1 a i, a i+1 a i... a d a i ] and therefore Proj(A) = d Spec ( ) d A (ai ) = i=1 i=1 These are the affine charts of Proj(A). ( [ a1 Spec C,..., a i 1, a i+1... a ]) d. a i a i a i a i Definition An affine monoid is a subset M Z d for some d Z 0 with 0 M and m 1, m 2 M = m 1 + m 2 M, and admitting a finitely generated subset {m 1, m 2,..., m n M, i.e., for all m, m = a 1 m a n m n for some a 1, a 2,..., a n Z 0. The affine monoid M is positive if the cone R + M R d is pointed. Proposition 3.2. A positive affine monoid has a smallest generating set, consisting of the indecomposible elements of M. These are elements that cannot be made from combinations of other elements of M. These elements are the Hilbert Basis, Hilb(M) of M. Hilb(M)={m M m 0 m = m 1 + m 2 for some m 1, m 2 M = m 1 = 0 or m 2 = 0
38 29 Theorem 3.3. Let A Z d be a finite set. Recall, C[A] = C[M] where M Z d+1 is the submonoid generated by {(a, 1) a A. Assume P=conv(A) R d. The following are equivalent 1. Proj(C[A]) = Proj(C[C(P ) Z d+1 ]) 2. For every v vert(p ), Hilb(R + (P v)) A v 3. C(P ) Z d+1 \ M is a finite set Proof. (1) (2) Let A w denote A w for all w V, where V = vert(p ) and P = conv(a). Similarly we shall let P w = C(P w) for all w V, the corner cones of P. Proj(C[M]) is covered by the affine varieties Spec(C[M w ]) for all w V, where M w denotes Z + (A w ). Let N denote C(P 1) Z d+1 and N w = P w Z d for each w V. Then likewise Proj(C[N]) is covered by Spec(C[N w ]) for all w V. Proj(C[M]) = Proj(C[N]) if and only if the affine varieties that cover each Proj are equal and that holds if and only if the rings C[M w ] and C[N w ] are equal. This will hold if and only if Hilb(R + (P v)) A v is an equality. Lemma 3.4. If (x, n) N and w V then there exists m 0 Z + such that m m 0 = (x, n) + m(w, 1) M. Proof. Since (x, n) N then (x, n) = v V q v(v, 1) where q v Q +. This implies that x = v V q vv and v V q v = n. Therefore x nw = v V q v (v w) N w
39 30 Since M w = N w, x nw = a A α a(a w), where α a Z +. Therefore in Z d+1, (x, n) = a A α a (a, 1) + (n a A α a )(w, 1) (x, n) + m(w, 1) = a A α a (a, 1) + (n a A α a + m)(w, 1). Therefore (x, n) + m(w, 1) M if m a A α a n. Let B = {y N : y = v V q vv where 0 q v < 1, then B is finite. By Lemma 1 there exists an integer m 0 such that m m 0 implies b + m(w, 1) M for all b B. Suppose y = v V q vv N and there exists w V such that q w m 0. Then y M. Proof. (3) = (2) Fix w V. Suppose x N w. Then x = v V q v(v w). For m sufficiently large we have (x + mw, m) = q v (v, 1) + (m q v )(w, 1) N v V v V Note that the two summands are in C(P {1) so their sum is in C(P {1) Z d+1 = N. The two individual summands are not necessarily in N. Since N \ M is a finite set there exists m Z + such that (x + mw, m) M. Therefore (x + mw, m) a A α a(a, 1) which implies x+mw = a A α aa and m = a A α a. Thus x = a A α a(a w) M w.
40 31 Figure 3.1: The gap behavior of a very ample polytope P. Definition Let A Z d be a finite set and P = conv(a) R d. The polytope P is very ample if for every v vert(p ), Hilb(R + (P v)) A v. Definition A polytope P R d is normal if vert(p ) Z d and for all c N, z (cp ) Z d there exists x 1, x 2,..., x c P Z d such that z = x 1 + x x c. It is known that all normal polytopes are very ample [3]. In [10] it is proved that there are very ample polytopes of dimension n 3 such that the n 2 multiples of them are not normal. This implies that there exist very ample polytopes that generate a non-zero finite amount of gaps. The natural question to ask is where do these gaps exist. In the rest of this chapter we will explore these non-normal very ample polytopes and their gaps.
41 Non-Normal Very Ample Polytope of Dimension d With h Gaps In [8] Higashitani showed that for given integers h 1 and d 3, there exists a non-normal very ample polytope of dimension d with exactly h gaps: let 0 if i = 1, e d if i = 2, e e d 1 if i = 3, h(e e d 1 + e d ) if i = 4, (h 1)(e e d 1 ) + he d if i = 5, u i = h(e e d 1 ) + (h 1)e d if i = 6, e 1 + (h + 1)e d if i = 7, e 1 + (h + 2)e d if i = 8, e 1 + e e d 1 + (h 3)e d if i = 9 e 1 + e e d 1 + (h 2)e d if i = 10, v i = e i, i = 2,..., d 1, v i = e i + e d, i = 2,..., d 1,
42 33 where 0 = (0, 0, 0,..., 0) R d and e 1, e 2,..., e d are the unit vectors of R d. The convex hull of {u 1,..., u 1 0 {v i, v i : i = 2,..., d 1 is a non-normal very ample polytope in dimension d with exactly h gaps. For this polytope the set of gaps is {(u j, 2) Z d+1 : j = 1,..., h, where u j = e 1 + j(e e d 1 ) + (j + h 1)e d. Let s see an example of a polytope P in dimension six with ten gaps, d = 6, h = 10: 0 if i = 1, (0, 0, 0, 0, 0, 1) if i = 2, (0, 1, 1, 1, 1, 0) if i = 3, (0, 10, 10, 10, 10, 10) if i = 4, (0, 9, 9, 9, 9, 10) if i = 5, u i = (0, 10, 10, 10, 10, 9) if i = 6, (1, 0, 0, 0, 0, 11) if i = 7, (1, 0, 0, 0, 0, 12) if i = 8, (1, 1, 1, 1, 1, 7) if i = 9 (1, 1, 1, 1, 1, 8) if i = 10,
43 34 v 1 = (0, 1, 0, 0, 0, 0) v 1 = (0, 1, 0, 0, 0, 1) v 2 = (0, 0, 1, 0, 0, 0) v 2 = (0, 0, 1, 0, 0, 1) v 3 = (0, 0, 0, 1, 0, 0) v 3 = (0, 0, 0, 1, 0, 1) v 4 = (0, 0, 0, 0, 1, 0) v 4 = (0, 0, 0, 0, 1, 1). The convex hull of all of these points forms our polytope P. Using Normaliz [2], a software package that computes the Hilbert basis of coned polytopes, one can input an integral polytope and the software will output the Hilbert basis of the cone generated by that polytope. Looking at the Hilbert basis of the cone, any elements that are not on height one are gaps: the Hilbert basis consists of all the indecomposable elements of the cone, therefore all elements on height one are included. If an element is in the Hilbert basis and is not on height one, it can not be reached with positive integer linear combinations of the height-one elements and by definition that makes it a gap. If the Hilbert basis is completely contained in height one the polytope is normal and there will be no gaps. According to Normaliz the
44 35 Hilbert basis elements not on height one for P are: (1, 1, 1, 1, 1, 10, 2) (1, 7, 7, 7, 7, 16, 2) (1, 3, 3, 3, 3, 12, 2) (1, 8, 8, 8, 8, 17, 2) (1, 4, 4, 4, 4, 13, 2) (1, 9, 9, 9, 9, 18, 2) (1, 5, 5, 5, 5, 14, 2) (1, 10, 10, 10, 10, 19, 2) (1, 6, 6, 6, 6, 15, 2) (1, 2, 2, 2, 2, 11, 2) As we can see we have exactly ten gaps and, as Higashitani claimed, all the gaps are on height two and are the elements of the set {(u j, 2) Z 7 : j = 1,..., 10, where u j = e 1 + j(e e 5 ) + (j + 9)e 6. Higashitani was able to create a non-normal very ample polytope in dimension d with h many gaps, however, there are no results about the gaps of an arbitrary non-normal very ample polytope. 3.3 Polytopal Affine Semigroups With Gaps Deep Inside Katthän constructed a non-normal very ample polytope P such that each gap has lattice distance at least k above every facet of P, where k is a positive integer [9]. Lattice distance between two lattice points x and y is the number of lattice points contained in conv(x, y) 1. Katthäns s polytope is a 3-simplex that is a special case
45 36 of a rectangular simplex. The construction of a rectangular simplex is as follows: Let λ = (λ 1, λ 2,..., λ n ) a vector of positive integers. Let = (λ) R n+1 with vertices v 0 := (0, 0,..., 0, 1) v 1 := (λ 1, 0, 0,..., 0, 1) v 2 := (0, λ 2, 0,..., 0, 1). v n := (0, 0,..., 0, λ n, 1) The polytope we are using is a 3-simplex and we will denote it as P (λ 1, λ 2, λ 3 ). As one would guess, not every λ i Z + will give us our desired result, Katthän describes the conditions, which he refers to as a good triple, for the λ i as λ 1 5 is odd, λ 2 = 2λ 1 1 and λ 3 = 2λ 2 1 λ 1 2. Katthan also goes on to prove that if P (λ 1, λ 2, λ 3 ) is a good triple then every gap has lattice distance at least λ above the closest facet. Lets see an example of this. Let λ 1 = 7 then our good triple will be (7, 13, 89) and the vertices of the polytope
46 37 are: v 0 = (0, 0, 0, 1) v 1 = (7, 0, 0, 1) v 2 = (0, 13, 0, 1) v 3 = (0, 0, 89, 1) Using Normaliz we get{(5, 11, 39, 2), (5, 12, 32, 2), (6, 9, 40, 2), (6, 10, 33, 2), (6, 11, 26, 2), (6, 12, 19, 2) as the set of gaps. 3.4 How to find gaps of very ample polytopes In the above papers the common theme was constructing very ample polytopes such that the gaps meet certain criteria. In [9] Katthän showed a construction of a very ample polytope that has gaps deep inside and in [8] Higashitani constructed a very ample polytope in dimension d with h gaps. Neither of these papers demonstrated a method to find the gaps of a general very ample polytope, which is what we will be discussing next. In order to compute the gaps of a very ample polytope we must first start with one. Given an arbitrary set of vertices, Normaliz can check each of the corner cones, also known as tangent cones, and verify whether or not the Hilbert basis of the cone
47 38 is contained in the polytope. Definition For a lattice polytope P Z d and v vert(p ) a corner cone of P is R + (P v). If the Hilbert Basis of every corner cone is contained within the polytope then P is very ample. As will be shown later, all the polytopes that were computed for this research were already known to be very ample. The best way to explain how to find the gaps of a very ample polytope is by example. The set of vertices {(0, 0, 0), (0, 0, 1), (0, 1, 0), (0, 1, 1), (1, 1, 0), (1, 1, 1), (1, 0, 5), (1, 0, 6) forms a very ample polytope P in dimension three. The following steps show how to find the gaps of P : 1. Input the vertices of P into Normaliz under the polytope mode
48 polytope 2. Record the Hilbert basis elements of degree Input these elements into the R formatting program to return the proper format for Python. 4. The elements will now be in a nested list format and can be pasted into Python as level 1. level1=[[1,0,6,1],[1,0,5,1],[1,1,1,1],[1,1,0,1],[0,1,1,1], [0,1,0,1],[0,0,1,1],[0,0,0,1]] 5. Use the addlevel function to return all elements of Z + (P, 1) on level 2, record output as replevel2. addlevel(level1,level1) 6. Use the scalarmult function to return a dilation of P in Normaliz format. scalarmult(2,level1) 7. Copy and paste this output into Normaliz to obtain the degree 1 elements.
49 40 8. Copy and paste these degree 1 elements into the R formatting program to return proper Python format. level2=[[0,0,0,2],[0,0,2,2],[0,2,0,2],[0,2,2,2],[2,2,0,2], [2,2,2,2],[2,0,10,2],[2,0,12,2],[0,0,1,2],[0,1,0,2],[0,1,1,2], [0,1,2,2],[0,2,1,2],[1,0,5,2],[1,0,6,2],[1,0,7,2],[1,1,0,2], [1,1,1,2],[1,1,2,2],[1,1,3,2],[1,1,4,2],[1,1,5,2],[1,1,6,2], [1,1,7,2],[1,2,0,2],[1,2,1,2],[1,2,2,2],[2,0,11,2],[2,1,5,2], [2,1,6,2],[2,1,7,2],[2,2,1,2]] 9. Copy and paste this properly formatted data for R + (P, 1) on level 2 into Python then use the gaps function to compare the lists of R + (P, 1) and Z + (P, 1) on level 2. What is returned will be the gaps on level 2. gaps(level2,replevel2) gaps on level2=[1, 1, 3, 2],[1, 1, 4, 2] 10. Continue this process until the gaps function returns zero gaps. 3.5 Bruns Polytopes In this section we will discuss a special variety of very ample polytopes constructed, and kindly provided, by Winfried Bruns. Bruns constructed this list of very ample polytopes by taking normal polytopes and, at random, removing elements contained
50 41 Table 3.1: Gaps of Bruns polytopes. Number of generators First height of no gaps gap vector 8 3 (0,1,0) 9 4 (0,2,3,0) 10 5 (0,3,6,3,0) 11 5 (0,4,6,6,0) 12 4 (0,2,3,0) 13 3 (0,1,0) 14 6 (0,7,12,12,10,0) 15 6 (0,7,12,12,10,0) 17 4 (0,4,3,0) 18 3 (0,1,0) 19 7 (0,14,30,42,40,30,0) 31 3 (0,1,0) in the polytope s corner cones. He compiled a list of hundreds of very ample polytopes in dimensions three, four, five and six. Table 3.1 shows the dimension, number of vertices and the first gap-free height of some of Bruns polytopes. 3.6 Very ample polytopes in dimension 3 Besides the polytopes in Bruns list, the other family of very ample polytopes that we will focus on are the Bruns and Gubeladze (BG) polytopes. Let I i, i = 1, 2, 3, 4 be intervals in Z, each containing at least two integers. The BG polytope is defined
51 42 as the convex hull of ((0, 0) I 1 ) ((0, 1) I 2 ) ((1, 1) I 3 ) ((1, 0) I 4 ). Figure 3.1 shows an example of a BG polytope. This class of polytopes can be found Figure 3.2: I 1 = {0, 1 I 2 = {0, 1 I 3 = {0, 1 I 4 = {4, 5 in [3]. Theorem 3.5. The class of BG polytopes are very ample. Proof. Consider the subgroup Γ of the group of affine isometries of R 3 that respect the integer lattice Z d, generated by the following three isometries: the rotation by 90 around the axis ( 1, 1, R), 2 2 (x, y, z) (x, y, z),
52 43 (x, y, z) (x, y, z + 1). For every vertex v P (I 1, I 2, I 3, I 4 ) there exists g Γ such that g(v) = 0. Moreover, every element of Γ permutes the set of lines (0, 0, R), (1, 0, R), (0, 1, R), (1, 1, R) R 3 Based on these observations, without loss of generality we can assume I 1 = [0, b 1 ] and we only need to check that C = R + P (I 1, I 2, I 3, I 4 ) Z 3 = Z + (1, 0, a 2 ) + Z + (0, 1, a 3 ) + Z + (1, 1, a 4 ) + Z + e, (3.1) where C = R + P (I 1, I 2, I 3, I 4 ) and e = (0, 0, 1). There are two possibilities: either C = R + e + R + (1, 0, a 2 ) + R + (0, 1, a 3 ), or C = ( R + e + R + (1, 0, a 2 ) + R + (1, 1, a 4 ) ) ( R + e + R + (0, 1, a 3 ) + R + (1, 1, a 4 ) ). In the first case, (3.1) follows from the observation that {e, (1, 0, a 2 ), (0, 1, a 3 ) is a basis of Z 3 : 1 0 a 2 det 0 1 a 3 = In the second case, (3.1) follows from the observation that the two cones on the
53 44 right-hand side are spanned by bases of Z 3 : 1 0 a 2 det 1 1 a 4 = 1, a 3 det 1 1 a 4 = 1, and, therefore, every lattice point in C is a non-negative integer linear combination of either {(1, 0, a 2 ), (1, 1, a 4 ), e, or {(0, 1, a 3 ), (1, 1, a 4 ), e. Table 3.2 shows the gaps of some of these polytopes for specific I i. Table 3.2: Gap heights of various BG polytopes. I 1 I 2 I 3 I 4 First height of no gaps {0, 1 {0, 1 {0, 1 {5, 6 4 {0, 1 {0, 1 {0, 1 {6, 7 5 {0, 1 {3, 4 {1, 2 {5, 6 6 {0, 1 {3, 4 {1, 2 {7, 8 8 {0, 1 {3, 4 {1, 2 {8, 9 9 {0, 2 {0, 2 {0, 2 {0, 2 2 {0, 5 {7, 10 {2, 5 {3, 7 2 {0, 1 {2, 3 {1, 2 {3, 4 3 {0, 1 {5, 6 {1, 2 {7, 8 10 {0, 1 {4, 5 {6, 7 {8, Is there a universal bound for the gaps? A natural question to ask, is there a universal bound for the gaps of all very ample polytopes of a certain dimension? Using the BG polytope construction for dimension
54 45 three we will show that there is not. In order to accomplish this we are going to introduce Ehrhart polynomials. Ehrhart polynomials are used to count lattice points within integral polytopes. In our case we are going to use them to count the lattice points at each height of R + (P, 1) where P is a BG polytope. Definition Let P be a polytope in dimension d and tp be an integer dilation of P. Then L(P, t) is the number of lattice points in tp. Moreover L(P, t) is a polynomial of the form a d t d + a d 1 t d a 0 where a d, a d 1,..., a 0 are rational [1]. This polynomial is known as the Ehrhart polynomial for a polytope P. It is known that a d is the volume of the polytope, a d 1 is half of the surface area of P and a 0 = 1 [1]. Theorem 3.6. A universal upper bound for the degrees of gaps of all very ample polytopes of dimension three does not exist. Proof. We will use the BG polytope construction for this proof. As a reminder, the construction goes as follows: Let I i, i = 1, 2, 3, 4 be intervals in Z, each containing at least two integers. The BG polytope is defined as the convex hull of ((0, 0) I 1 ) ((0, 1) I 2 ) ((1, 1) I 3 ) ((1, 0) I 4 ). For this proof we will let I 1 = I 2 = I 3 = {0, 1 and I 4 = {k, k +1 where k Z 0 and denote this polytope by P k. An example of this polytope can be seen in Figure
55 The volume of P k is k k 6 = k 6 + 1; furthermore, P k has four triangular facets and four square facets all with empty interiors, which implies the surface area is 4( 1 2 (1)) + 4(1)) = 6. Therefore L(P k, t) = ( k 6 + 1)t3 + 3t 2 + l k t + 1, we also know L(P k, 1) = 8 which implies l k = 3 k 6, thus L(P k, t) = ( ) ( k t 3 + 3t k ) t Note that for t 2, L(P k, t) k 6 t3 k 6 t k 2 (3.2) Assume by contradiction that there exists a bound b Z + such that for all k γ(p k ) b (3.3) where γ(p k ) Z is the largest height containing gaps. The underlying point configuration (P k, 1) Z 4 is (P k, 1) = k k + 1.
56 47 Since P k is very ample for t > γ(p k ), dim K (K(Z + (P k, 1)) at height t is equal to L(P k, t) 8 t. Using the 8 vertices of P k, for each height t there is 8 t ways to sum t-many of the 8 vertices where repeats are allowed. Of course many combinations may sum to the same amount which is why L(P k, t) 8 t. Combining this with (3.2) and (3.3) gives for t > b k 2 8t which is a contradiction since k can be arbitrarily large. Notice that the last non-zero index of each gap vector of Table 3.3 is the (k 3)rd tetrahedral number. So far we have not been able to explain why this is the case. Also we have not been able to explain why every diagonal is a multiple of its corresponding tetrahedral number. Table 3.3: Gaps of BG polytope when I 1 = I 2 = I 3 = {0, 1 and I 4 = {k, k + 1. k First height of no gaps Gap vector Gaps 4 3 (0,1,0) (0,2,4,0) (0,3,8,10,0) (0,4,12,20,20,0) (0,5,16,30,40, 35,0) (0,6,20,40,60,70,56,0) (0,7,24,50,80,105,112,84,0) (0,8,28,60,100,140,168,168,120,0) (0,9,32,70,120,175,224,252,240,165,0) (0,10,36,80,140,210,280,336,360,330,220,0) 2002 Conjecture 3.1. For any arbitrary k let G k,i denote the i th index of the gap vector
57 48 associated with k. We conjecture that ( ) i + 1 G k,i = (k i 1) i = 1, 2,..., k 1 3 with k 1 being the first height of no gaps.
58 Appendix A: JAVA code to generate gaps for a given a import java.util.scanner; public class FastFrobenius { import java.util.scanner; public static void main(string[] args) { Scanner scanner = new Scanner( System.in ); System.out.println("enter a positive integer"); 49
59 50 double a = scanner.nextdouble(); double bound=1/(a+1); double b=math.pow(a,3); double[][] pq = new double[(int) (b+1)][2]; for (int i=0; i<=b;i++){ pq[i][0] = (int) (i%(a*a)+1); pq[i][1] = (int) (((i%(a*a))-i)/(-(a*a))+1); if( (pq[i][1]/pq[i][0])<bound ){ pq[i][0]=0; pq[i][1]=0; for (int i=0; i<=b;i++){ double a2mod= i%(a*a); double w1=(int) (a2mod+1); double w2=(int) ((a2mod-i)/(-(a*a))+1);
60 51 if( (w2/w1)>=bound ){ pq[i][0] = w1; pq[i][1] = w2; System.out.println("OK"); for (int i=0; i<=b;i++){ for (int y=0; y<=2*a;y++){ for(int z=0;z<=2*a;z++){ if( (pq[i][0]-(a*y)-(z*a)-z>=0) && (pq[i][0]-(a*y)-(z*a)-z<=2*a) && (pq[i][1]-(pq[i][0]-(a*y)-(z*a)-z)-y-z>=0) && (pq[i][1]-(pq[i][0]-(a*y)-(z*a)-z)-y-z<=2*a) ){ pq[i][0]=0; pq[i][1]=0;
61 52 for (int r=0; r<pq.length; r++) { for (int c=0; c<pq[r].length; c++) { System.out.print((int) pq[r][0]+","); System.out.println("");
62 Appendix B: Python code that computes gaps of very ample polytopes def getrow(mat, row): "return row row from matrix mat" return(mat[row]) def cols(mat): "return number of cols" return(len(mat[0])) def addvectors(a,b): "add two vectors" if len(a)!= len(b): print("addvectors: different lengths") 53
63 54 return() return([a[i]+b[i] for i in range(len(a))]) def rows(mat): "return number of rows" return(len(mat)) def addlevel(a,b): "Adds each vector to each vector" L= [0]*(rows(A)*rows(B)) for i in range(rows(a)): for j in range (rows(b)): "print (addvectors(a[i],b[j]))" L[(rows(B)*i)+j]=addVectors(A[i],B[j]) duplicates(l) L.remove(0) print (L) def gaps (A,B): "A is big list B is small list"
64 55 for i in range(rows(a)): if A[i] not in B: print (A[i]) def duplicates (mat): "deletes duplicate vectors" for i in range(rows(mat)): for j in range (rows(mat)): if i== j: continue else: if mat[i]==mat[j]: mat[i]=0 def scalarmult(a,mat): "multiplies a scalar times a matrix" "If mat is a vector it returns a vector." if vectorq(mat):
65 56 return([a*m for m in mat]) for row in range(rows(mat)): for col in range(cols(mat)): mat[row][col] = a*mat[row][col] return(mat) def vectorq(v): "mild test to see if V is a vector" if type(v)!= type([1]): return(false) if type(v[0]) == type([1]): return(false) return(true)
66 Appendix C: R formatting program myformat=function(normaliz){ n=nrow(normaliz) x="[" for(i in 1:(n-1)){ x=paste(x,"[",normaliz[i,1],",",normaliz[i,2],",",normaliz[i,3],",",normaliz[i,4],"],",sep="") x=paste(x,"[",normaliz[n,1],",",normaliz[n,2],",",normaliz[n,3],",",normaliz[n,4],"]]",sep="") return(x) myformat2=function(normaliz){ n=nrow(normaliz) 57
67 58 x="[" for(i in 1:(n-1)){ x=paste(x,"[",normaliz[i,1],",",normaliz[i,2],",",normaliz[i,3],",","2","],",sep="") x=paste(x,"[",normaliz[n,1],",",normaliz[n,2],",",normaliz[n,3],",","2","]]",sep="") return(x) myformat3=function(normaliz){ n=nrow(normaliz) x="[" for(i in 1:(n-1)){ x=paste(x,"[",normaliz[i,1],",",normaliz[i,2],",",normaliz[i,3],",","3","],",sep="") x=paste(x,"[",normaliz[n,1],",",normaliz[n,2],",",normaliz[n,3],",","3","]]",sep="")
68 59 return(x) myformat4=function(normaliz){ n=nrow(normaliz) x="[" for(i in 1:(n-1)){ x=paste(x,"[",normaliz[i,1],",",normaliz[i,2],",",normaliz[i,3],",","4","],",sep="") x=paste(x,"[",normaliz[n,1],",",normaliz[n,2],",",normaliz[n,3],",","4","]]",sep="") return(x) myformat5=function(normaliz){ n=nrow(normaliz) x="[" for(i in 1:(n-1)){ x=paste(x,"[",normaliz[i,1],",",normaliz[i,2],","
69 60,normaliz[i,3],",","5","],",sep="") x=paste(x,"[",normaliz[n,1],",",normaliz[n,2],",",normaliz[n,3],",","5","]]",sep="") return(x) myformat6=function(normaliz){ n=nrow(normaliz) x="[" for(i in 1:(n-1)){ x=paste(x,"[",normaliz[i,1],",",normaliz[i,2],",",normaliz[i,3],",","6","],",sep="") x=paste(x,"[",normaliz[n,1],",",normaliz[n,2],",",normaliz[n,3],",","6","]]",sep="") return(x)
Classification of Ehrhart quasi-polynomials of half-integral polygons
Classification of Ehrhart quasi-polynomials of half-integral polygons A thesis presented to the faculty of San Francisco State University In partial fulfilment of The Requirements for The Degree Master
More informationMATH 890 HOMEWORK 2 DAVID MEREDITH
MATH 890 HOMEWORK 2 DAVID MEREDITH (1) Suppose P and Q are polyhedra. Then P Q is a polyhedron. Moreover if P and Q are polytopes then P Q is a polytope. The facets of P Q are either F Q where F is a facet
More informationInterval-Vector Polytopes
Interval-Vector Polytopes Jessica De Silva Gabriel Dorfsman-Hopkins California State University, Stanislaus Dartmouth College Joseph Pruitt California State University, Long Beach July 28, 2012 Abstract
More informationLecture 2 - Introduction to Polytopes
Lecture 2 - Introduction to Polytopes Optimization and Approximation - ENS M1 Nicolas Bousquet 1 Reminder of Linear Algebra definitions Let x 1,..., x m be points in R n and λ 1,..., λ m be real numbers.
More informationNumerical Optimization
Convex Sets Computer Science and Automation Indian Institute of Science Bangalore 560 012, India. NPTEL Course on Let x 1, x 2 R n, x 1 x 2. Line and line segment Line passing through x 1 and x 2 : {y
More informationLecture 3. Corner Polyhedron, Intersection Cuts, Maximal Lattice-Free Convex Sets. Tepper School of Business Carnegie Mellon University, Pittsburgh
Lecture 3 Corner Polyhedron, Intersection Cuts, Maximal Lattice-Free Convex Sets Gérard Cornuéjols Tepper School of Business Carnegie Mellon University, Pittsburgh January 2016 Mixed Integer Linear Programming
More informationDM545 Linear and Integer Programming. Lecture 2. The Simplex Method. Marco Chiarandini
DM545 Linear and Integer Programming Lecture 2 The Marco Chiarandini Department of Mathematics & Computer Science University of Southern Denmark Outline 1. 2. 3. 4. Standard Form Basic Feasible Solutions
More informationCyclotomic Polytopes and Growth Series of Cyclotomic Lattices. Matthias Beck & Serkan Hoşten San Francisco State University
Cyclotomic Polytopes and Growth Series of Cyclotomic Lattices Matthias Beck & Serkan Hoşten San Francisco State University math.sfsu.edu/beck Math Research Letters Growth Series of Lattices L R d lattice
More informationmaximize c, x subject to Ax b,
Lecture 8 Linear programming is about problems of the form maximize c, x subject to Ax b, where A R m n, x R n, c R n, and b R m, and the inequality sign means inequality in each row. The feasible set
More informationPOLYHEDRAL GEOMETRY. Convex functions and sets. Mathematical Programming Niels Lauritzen Recall that a subset C R n is convex if
POLYHEDRAL GEOMETRY Mathematical Programming Niels Lauritzen 7.9.2007 Convex functions and sets Recall that a subset C R n is convex if {λx + (1 λ)y 0 λ 1} C for every x, y C and 0 λ 1. A function f :
More informationAdvanced Operations Research Techniques IE316. Quiz 1 Review. Dr. Ted Ralphs
Advanced Operations Research Techniques IE316 Quiz 1 Review Dr. Ted Ralphs IE316 Quiz 1 Review 1 Reading for The Quiz Material covered in detail in lecture. 1.1, 1.4, 2.1-2.6, 3.1-3.3, 3.5 Background material
More informationLECTURE 1 Basic definitions, the intersection poset and the characteristic polynomial
R. STANLEY, HYPERPLANE ARRANGEMENTS LECTURE Basic definitions, the intersection poset and the characteristic polynomial.. Basic definitions The following notation is used throughout for certain sets of
More informationarxiv: v1 [math.co] 25 Sep 2015
A BASIS FOR SLICING BIRKHOFF POLYTOPES TREVOR GLYNN arxiv:1509.07597v1 [math.co] 25 Sep 2015 Abstract. We present a change of basis that may allow more efficient calculation of the volumes of Birkhoff
More informationConvex Geometry arising in Optimization
Convex Geometry arising in Optimization Jesús A. De Loera University of California, Davis Berlin Mathematical School Summer 2015 WHAT IS THIS COURSE ABOUT? Combinatorial Convexity and Optimization PLAN
More informationEC 521 MATHEMATICAL METHODS FOR ECONOMICS. Lecture 2: Convex Sets
EC 51 MATHEMATICAL METHODS FOR ECONOMICS Lecture : Convex Sets Murat YILMAZ Boğaziçi University In this section, we focus on convex sets, separating hyperplane theorems and Farkas Lemma. And as an application
More informationACTUALLY DOING IT : an Introduction to Polyhedral Computation
ACTUALLY DOING IT : an Introduction to Polyhedral Computation Jesús A. De Loera Department of Mathematics Univ. of California, Davis http://www.math.ucdavis.edu/ deloera/ 1 What is a Convex Polytope? 2
More informationMA4254: Discrete Optimization. Defeng Sun. Department of Mathematics National University of Singapore Office: S Telephone:
MA4254: Discrete Optimization Defeng Sun Department of Mathematics National University of Singapore Office: S14-04-25 Telephone: 6516 3343 Aims/Objectives: Discrete optimization deals with problems of
More informationCombinatorial Geometry & Topology arising in Game Theory and Optimization
Combinatorial Geometry & Topology arising in Game Theory and Optimization Jesús A. De Loera University of California, Davis LAST EPISODE... We discuss the content of the course... Convex Sets A set is
More informationAMS : Combinatorial Optimization Homework Problems - Week V
AMS 553.766: Combinatorial Optimization Homework Problems - Week V For the following problems, A R m n will be m n matrices, and b R m. An affine subspace is the set of solutions to a a system of linear
More informationOn the undecidability of the tiling problem. Jarkko Kari. Mathematics Department, University of Turku, Finland
On the undecidability of the tiling problem Jarkko Kari Mathematics Department, University of Turku, Finland Consider the following decision problem, the tiling problem: Given a finite set of tiles (say,
More informationCharacterizing Automorphism and Permutation Polytopes
Characterizing Automorphism and Permutation Polytopes By KATHERINE EMILIE JONES BURGGRAF SENIOR THESIS Submitted in partial satisfaction of the requirements for Highest Honors for the degree of BACHELOR
More informationLecture notes for Topology MMA100
Lecture notes for Topology MMA100 J A S, S-11 1 Simplicial Complexes 1.1 Affine independence A collection of points v 0, v 1,..., v n in some Euclidean space R N are affinely independent if the (affine
More informationHowever, this is not always true! For example, this fails if both A and B are closed and unbounded (find an example).
98 CHAPTER 3. PROPERTIES OF CONVEX SETS: A GLIMPSE 3.2 Separation Theorems It seems intuitively rather obvious that if A and B are two nonempty disjoint convex sets in A 2, then there is a line, H, separating
More informationarxiv: v1 [math.co] 12 Dec 2017
arxiv:1712.04381v1 [math.co] 12 Dec 2017 Semi-reflexive polytopes Tiago Royer Abstract The Ehrhart function L P(t) of a polytope P is usually defined only for integer dilation arguments t. By allowing
More informationLecture 2. Topology of Sets in R n. August 27, 2008
Lecture 2 Topology of Sets in R n August 27, 2008 Outline Vectors, Matrices, Norms, Convergence Open and Closed Sets Special Sets: Subspace, Affine Set, Cone, Convex Set Special Convex Sets: Hyperplane,
More information4. Simplicial Complexes and Simplicial Homology
MATH41071/MATH61071 Algebraic topology Autumn Semester 2017 2018 4. Simplicial Complexes and Simplicial Homology Geometric simplicial complexes 4.1 Definition. A finite subset { v 0, v 1,..., v r } R n
More informationClassification of smooth Fano polytopes
Classification of smooth Fano polytopes Mikkel Øbro PhD thesis Advisor: Johan Peder Hansen September 2007 Department of Mathematical Sciences, Faculty of Science University of Aarhus Introduction This
More informationFACES OF CONVEX SETS
FACES OF CONVEX SETS VERA ROSHCHINA Abstract. We remind the basic definitions of faces of convex sets and their basic properties. For more details see the classic references [1, 2] and [4] for polytopes.
More informationOrientation of manifolds - definition*
Bulletin of the Manifold Atlas - definition (2013) Orientation of manifolds - definition* MATTHIAS KRECK 1. Zero dimensional manifolds For zero dimensional manifolds an orientation is a map from the manifold
More information3. The Simplex algorithmn The Simplex algorithmn 3.1 Forms of linear programs
11 3.1 Forms of linear programs... 12 3.2 Basic feasible solutions... 13 3.3 The geometry of linear programs... 14 3.4 Local search among basic feasible solutions... 15 3.5 Organization in tableaus...
More informationInteger Programming Theory
Integer Programming Theory Laura Galli October 24, 2016 In the following we assume all functions are linear, hence we often drop the term linear. In discrete optimization, we seek to find a solution x
More informationAutomorphism Groups of Cyclic Polytopes
8 Automorphism Groups of Cyclic Polytopes (Volker Kaibel and Arnold Waßmer ) It is probably well-known to most polytope theorists that the combinatorial automorphism group of a cyclic d-polytope with n
More informationA combinatorial proof of a formula for Betti numbers of a stacked polytope
A combinatorial proof of a formula for Betti numbers of a staced polytope Suyoung Choi Department of Mathematical Sciences KAIST, Republic of Korea choisy@aistacr (Current Department of Mathematics Osaa
More informationOn Soft Topological Linear Spaces
Republic of Iraq Ministry of Higher Education and Scientific Research University of AL-Qadisiyah College of Computer Science and Formation Technology Department of Mathematics On Soft Topological Linear
More informationConic Duality. yyye
Conic Linear Optimization and Appl. MS&E314 Lecture Note #02 1 Conic Duality Yinyu Ye Department of Management Science and Engineering Stanford University Stanford, CA 94305, U.S.A. http://www.stanford.edu/
More informationInstitutionen för matematik, KTH.
Institutionen för matematik, KTH. Chapter 10 projective toric varieties and polytopes: definitions 10.1 Introduction Tori varieties are algebraic varieties related to the study of sparse polynomials.
More informationLinear Programming in Small Dimensions
Linear Programming in Small Dimensions Lekcija 7 sergio.cabello@fmf.uni-lj.si FMF Univerza v Ljubljani Edited from slides by Antoine Vigneron Outline linear programming, motivation and definition one dimensional
More informationTORIC VARIETIES JOAQUÍN MORAGA
TORIC VARIETIES Abstract. This is a very short introduction to some concepts around toric varieties, some of the subsections are intended for more experienced algebraic geometers. To see a lot of exercises
More informationTechnische Universität München Zentrum Mathematik
Technische Universität München Zentrum Mathematik Prof. Dr. Dr. Jürgen Richter-Gebert, Bernhard Werner Projective Geometry SS 208 https://www-m0.ma.tum.de/bin/view/lehre/ss8/pgss8/webhome Solutions for
More informationMonotone Paths in Geometric Triangulations
Monotone Paths in Geometric Triangulations Adrian Dumitrescu Ritankar Mandal Csaba D. Tóth November 19, 2017 Abstract (I) We prove that the (maximum) number of monotone paths in a geometric triangulation
More informationSTANLEY S SIMPLICIAL POSET CONJECTURE, AFTER M. MASUDA
Communications in Algebra, 34: 1049 1053, 2006 Copyright Taylor & Francis Group, LLC ISSN: 0092-7872 print/1532-4125 online DOI: 10.1080/00927870500442005 STANLEY S SIMPLICIAL POSET CONJECTURE, AFTER M.
More informationPolytopes Course Notes
Polytopes Course Notes Carl W. Lee Department of Mathematics University of Kentucky Lexington, KY 40506 lee@ms.uky.edu Fall 2013 i Contents 1 Polytopes 1 1.1 Convex Combinations and V-Polytopes.....................
More informationChapter 4 Concepts from Geometry
Chapter 4 Concepts from Geometry An Introduction to Optimization Spring, 2014 Wei-Ta Chu 1 Line Segments The line segment between two points and in R n is the set of points on the straight line joining
More informationLattice-point generating functions for free sums of polytopes
Lattice-point generating functions for free sums of polytopes Tyrrell B. McAllister 1 Joint work with Matthias Beck 2 and Pallavi Jayawant 3 1 University of Wyoming 2 San Francisco State University 3 Bates
More informationLecture notes on the simplex method September We will present an algorithm to solve linear programs of the form. maximize.
Cornell University, Fall 2017 CS 6820: Algorithms Lecture notes on the simplex method September 2017 1 The Simplex Method We will present an algorithm to solve linear programs of the form maximize subject
More informationIndependence Complexes of Certain Families of Graphs
Independence Complexes of Certain Families of Graphs Rickard Fors rifo0@kth.se Master thesis in Mathematics at KTH Presented August 19, 011 Supervisor: Jakob Jonsson Abstract The focus of this report
More information2 The Fractional Chromatic Gap
C 1 11 2 The Fractional Chromatic Gap As previously noted, for any finite graph. This result follows from the strong duality of linear programs. Since there is no such duality result for infinite linear
More informationThe Subdivision of Large Simplicial Cones in Normaliz ICMS 2016
ICMS 2016 The Subdivision of Large Simplicial Cones in Normaliz Richard Sieg joint with Winfried Bruns & Christof Söger NORMALIZ NORMALIZ Open source software (GPL) written in C++ (using Boost and GMP/MPIR)
More informationSome Advanced Topics in Linear Programming
Some Advanced Topics in Linear Programming Matthew J. Saltzman July 2, 995 Connections with Algebra and Geometry In this section, we will explore how some of the ideas in linear programming, duality theory,
More informationBraid groups and Curvature Talk 2: The Pieces
Braid groups and Curvature Talk 2: The Pieces Jon McCammond UC Santa Barbara Regensburg, Germany Sept 2017 Rotations in Regensburg Subsets, Subdisks and Rotations Recall: for each A [n] of size k > 1 with
More informationMATH 423 Linear Algebra II Lecture 17: Reduced row echelon form (continued). Determinant of a matrix.
MATH 423 Linear Algebra II Lecture 17: Reduced row echelon form (continued). Determinant of a matrix. Row echelon form A matrix is said to be in the row echelon form if the leading entries shift to the
More informationMathematical and Algorithmic Foundations Linear Programming and Matchings
Adavnced Algorithms Lectures Mathematical and Algorithmic Foundations Linear Programming and Matchings Paul G. Spirakis Department of Computer Science University of Patras and Liverpool Paul G. Spirakis
More informationToric Varieties and Lattice Polytopes
Toric Varieties and Lattice Polytopes Ursula Whitcher April 1, 006 1 Introduction We will show how to construct spaces called toric varieties from lattice polytopes. Toric fibrations correspond to slices
More informationCREPANT RESOLUTIONS OF GORENSTEIN TORIC SINGULARITIES AND UPPER BOUND THEOREM. Dimitrios I. Dais
Séminaires & Congrès 6, 2002, p. 187 192 CREPANT RESOLUTIONS OF GORENSTEIN TORIC SINGULARITIES AND UPPER BOUND THEOREM by Dimitrios I. Dais Abstract. A necessary condition for the existence of torus-equivariant
More information2017 SOLUTIONS (PRELIMINARY VERSION)
SIMON MARAIS MATHEMATICS COMPETITION 07 SOLUTIONS (PRELIMINARY VERSION) This document will be updated to include alternative solutions provided by contestants, after the competition has been mared. Problem
More informationRigidity, connectivity and graph decompositions
First Prev Next Last Rigidity, connectivity and graph decompositions Brigitte Servatius Herman Servatius Worcester Polytechnic Institute Page 1 of 100 First Prev Next Last Page 2 of 100 We say that a framework
More informationLecture 0: Reivew of some basic material
Lecture 0: Reivew of some basic material September 12, 2018 1 Background material on the homotopy category We begin with the topological category TOP, whose objects are topological spaces and whose morphisms
More informationarxiv: v1 [math.gr] 19 Dec 2018
GROWTH SERIES OF CAT(0) CUBICAL COMPLEXES arxiv:1812.07755v1 [math.gr] 19 Dec 2018 BORIS OKUN AND RICHARD SCOTT Abstract. Let X be a CAT(0) cubical complex. The growth series of X at x is G x (t) = y V
More informationSPERNER S LEMMA MOOR XU
SPERNER S LEMMA MOOR XU Abstract. Is it possible to dissect a square into an odd number of triangles of equal area? This question was first answered by Paul Monsky in 970, and the solution requires elements
More informationMath 5593 Linear Programming Lecture Notes
Math 5593 Linear Programming Lecture Notes Unit II: Theory & Foundations (Convex Analysis) University of Colorado Denver, Fall 2013 Topics 1 Convex Sets 1 1.1 Basic Properties (Luenberger-Ye Appendix B.1).........................
More informationTriangulations Of Point Sets
Triangulations Of Point Sets Applications, Structures, Algorithms. Jesús A. De Loera Jörg Rambau Francisco Santos MSRI Summer school July 21 31, 2003 (Book under construction) Triangulations Of Point Sets
More informationTopological properties of convex sets
Division of the Humanities and Social Sciences Ec 181 KC Border Convex Analysis and Economic Theory Winter 2018 Topic 5: Topological properties of convex sets 5.1 Interior and closure of convex sets Let
More informationarxiv: v1 [math.co] 15 Dec 2009
ANOTHER PROOF OF THE FACT THAT POLYHEDRAL CONES ARE FINITELY GENERATED arxiv:092.2927v [math.co] 5 Dec 2009 VOLKER KAIBEL Abstract. In this note, we work out a simple inductive proof showing that every
More informationPebble Sets in Convex Polygons
2 1 Pebble Sets in Convex Polygons Kevin Iga, Randall Maddox June 15, 2005 Abstract Lukács and András posed the problem of showing the existence of a set of n 2 points in the interior of a convex n-gon
More informationIntersection Cuts with Infinite Split Rank
Intersection Cuts with Infinite Split Rank Amitabh Basu 1,2, Gérard Cornuéjols 1,3,4 François Margot 1,5 April 2010; revised April 2011; revised October 2011 Abstract We consider mixed integer linear programs
More informationChapter 18 out of 37 from Discrete Mathematics for Neophytes: Number Theory, Probability, Algorithms, and Other Stuff by J. M. Cargal.
Chapter 8 out of 7 from Discrete Mathematics for Neophytes: Number Theory, Probability, Algorithms, and Other Stuff by J. M. Cargal 8 Matrices Definitions and Basic Operations Matrix algebra is also known
More informationDefinition. Given a (v,k,λ)- BIBD, (X,B), a set of disjoint blocks of B which partition X is called a parallel class.
Resolvable BIBDs Definition Given a (v,k,λ)- BIBD, (X,B), a set of disjoint blocks of B which partition X is called a parallel class. A partition of B into parallel classes (there must be r of them) is
More informationarxiv: v2 [math.co] 4 Jul 2018
SMOOTH CENTRALLY SYMMETRIC POLYTOPES IN DIMENSION 3 ARE IDP MATTHIAS BECK, CHRISTIAN HAASE, AKIHIRO HIGASHITANI, JOHANNES HOFSCHEIER, KATHARINA JOCHEMKO, LUKAS KATTHÄN, AND MATEUSZ MICHAŁEK arxiv:802.0046v2
More informationNewton Polygons of L-Functions
Newton Polygons of L-Functions Phong Le Department of Mathematics University of California, Irvine June 2009/Ph.D. Defense Phong Le Newton Polygons of L-Functions 1/33 Laurent Polynomials Let q = p a where
More informationCS675: Convex and Combinatorial Optimization Spring 2018 Consequences of the Ellipsoid Algorithm. Instructor: Shaddin Dughmi
CS675: Convex and Combinatorial Optimization Spring 2018 Consequences of the Ellipsoid Algorithm Instructor: Shaddin Dughmi Outline 1 Recapping the Ellipsoid Method 2 Complexity of Convex Optimization
More informationCardinality of Sets. Washington University Math Circle 10/30/2016
Cardinality of Sets Washington University Math Circle 0/0/06 The cardinality of a finite set A is just the number of elements of A, denoted by A. For example, A = {a, b, c, d}, B = {n Z : n } = {,,, 0,,,
More informationLecture 6: Faces, Facets
IE 511: Integer Programming, Spring 2019 31 Jan, 2019 Lecturer: Karthik Chandrasekaran Lecture 6: Faces, Facets Scribe: Setareh Taki Disclaimer: These notes have not been subjected to the usual scrutiny
More information11.1 Facility Location
CS787: Advanced Algorithms Scribe: Amanda Burton, Leah Kluegel Lecturer: Shuchi Chawla Topic: Facility Location ctd., Linear Programming Date: October 8, 2007 Today we conclude the discussion of local
More informationConvexity: an introduction
Convexity: an introduction Geir Dahl CMA, Dept. of Mathematics and Dept. of Informatics University of Oslo 1 / 74 1. Introduction 1. Introduction what is convexity where does it arise main concepts and
More information2 Geometry Solutions
2 Geometry Solutions jacques@ucsd.edu Here is give problems and solutions in increasing order of difficulty. 2.1 Easier problems Problem 1. What is the minimum number of hyperplanar slices to make a d-dimensional
More informationFano varieties and polytopes
Fano varieties and polytopes Olivier DEBARRE The Fano Conference Torino, September 29 October 5, 2002 Smooth Fano varieties A smooth Fano variety (over a fixed algebraically closed field k) is a smooth
More informationOn Rainbow Cycles in Edge Colored Complete Graphs. S. Akbari, O. Etesami, H. Mahini, M. Mahmoody. Abstract
On Rainbow Cycles in Edge Colored Complete Graphs S. Akbari, O. Etesami, H. Mahini, M. Mahmoody Abstract In this paper we consider optimal edge colored complete graphs. We show that in any optimal edge
More informationRATIONAL CURVES ON SMOOTH CUBIC HYPERSURFACES. Contents 1. Introduction 1 2. The proof of Theorem References 9
RATIONAL CURVES ON SMOOTH CUBIC HYPERSURFACES IZZET COSKUN AND JASON STARR Abstract. We prove that the space of rational curves of a fixed degree on any smooth cubic hypersurface of dimension at least
More informationThe Simplex Algorithm
The Simplex Algorithm Uri Feige November 2011 1 The simplex algorithm The simplex algorithm was designed by Danzig in 1947. This write-up presents the main ideas involved. It is a slight update (mostly
More informationBasic Properties The Definition of Catalan Numbers
1 Basic Properties 1.1. The Definition of Catalan Numbers There are many equivalent ways to define Catalan numbers. In fact, the main focus of this monograph is the myriad combinatorial interpretations
More informationCHAPTER 5 SYSTEMS OF EQUATIONS. x y
page 1 of Section 5.1 CHAPTER 5 SYSTEMS OF EQUATIONS SECTION 5.1 GAUSSIAN ELIMINATION matrix form of a system of equations The system 2x + 3y + 4z 1 5x + y + 7z 2 can be written as Ax where b 2 3 4 A [
More informationLecture 2 September 3
EE 381V: Large Scale Optimization Fall 2012 Lecture 2 September 3 Lecturer: Caramanis & Sanghavi Scribe: Hongbo Si, Qiaoyang Ye 2.1 Overview of the last Lecture The focus of the last lecture was to give
More informationLecture 15: The subspace topology, Closed sets
Lecture 15: The subspace topology, Closed sets 1 The Subspace Topology Definition 1.1. Let (X, T) be a topological space with topology T. subset of X, the collection If Y is a T Y = {Y U U T} is a topology
More informationGlossary Common Core Curriculum Maps Math/Grade 6 Grade 8
Glossary Common Core Curriculum Maps Math/Grade 6 Grade 8 Grade 6 Grade 8 absolute value Distance of a number (x) from zero on a number line. Because absolute value represents distance, the absolute value
More informationPolyhedral Computation and their Applications. Jesús A. De Loera Univ. of California, Davis
Polyhedral Computation and their Applications Jesús A. De Loera Univ. of California, Davis 1 1 Introduction It is indeniable that convex polyhedral geometry is an important tool of modern mathematics.
More informationToric Cohomological Rigidity of Simple Convex Polytopes
Toric Cohomological Rigidity of Simple Convex Polytopes Dong Youp Suh (KAIST) The Second East Asian Conference on Algebraic Topology National University of Singapore December 15-19, 2008 1/ 28 This talk
More informationThe Construction of a Hyperbolic 4-Manifold with a Single Cusp, Following Kolpakov and Martelli. Christopher Abram
The Construction of a Hyperbolic 4-Manifold with a Single Cusp, Following Kolpakov and Martelli by Christopher Abram A Thesis Presented in Partial Fulfillment of the Requirement for the Degree Master of
More informationPACKING DIGRAPHS WITH DIRECTED CLOSED TRAILS
PACKING DIGRAPHS WITH DIRECTED CLOSED TRAILS PAUL BALISTER Abstract It has been shown [Balister, 2001] that if n is odd and m 1,, m t are integers with m i 3 and t i=1 m i = E(K n) then K n can be decomposed
More informationON THE STRONGLY REGULAR GRAPH OF PARAMETERS
ON THE STRONGLY REGULAR GRAPH OF PARAMETERS (99, 14, 1, 2) SUZY LOU AND MAX MURIN Abstract. In an attempt to find a strongly regular graph of parameters (99, 14, 1, 2) or to disprove its existence, we
More informationarxiv: v2 [math.co] 24 Aug 2016
Slicing and dicing polytopes arxiv:1608.05372v2 [math.co] 24 Aug 2016 Patrik Norén June 23, 2018 Abstract Using tropical convexity Dochtermann, Fink, and Sanyal proved that regular fine mixed subdivisions
More informationGeometric structures on manifolds
CHAPTER 3 Geometric structures on manifolds In this chapter, we give our first examples of hyperbolic manifolds, combining ideas from the previous two chapters. 3.1. Geometric structures 3.1.1. Introductory
More informationL-CONVEX-CONCAVE SETS IN REAL PROJECTIVE SPACE AND L-DUALITY
MOSCOW MATHEMATICAL JOURNAL Volume 3, Number 3, July September 2003, Pages 1013 1037 L-CONVEX-CONCAVE SETS IN REAL PROJECTIVE SPACE AND L-DUALITY A. KHOVANSKII AND D. NOVIKOV Dedicated to Vladimir Igorevich
More informationTechnische Universität München Zentrum Mathematik
Question 1. Incidence matrix with gaps Technische Universität München Zentrum Mathematik Prof. Dr. Dr. Jürgen Richter-Gebert, Bernhard Werner Projective Geometry SS 2016 www-m10.ma.tum.de/projektivegeometriess16
More informationCourse Number 432/433 Title Algebra II (A & B) H Grade # of Days 120
Whitman-Hanson Regional High School provides all students with a high- quality education in order to develop reflective, concerned citizens and contributing members of the global community. Course Number
More informationLinear Programming and its Applications
Linear Programming and its Applications Outline for Today What is linear programming (LP)? Examples Formal definition Geometric intuition Why is LP useful? A first look at LP algorithms Duality Linear
More informationarxiv: v1 [math.co] 27 Feb 2015
Mode Poset Probability Polytopes Guido Montúfar 1 and Johannes Rauh 2 arxiv:1503.00572v1 [math.co] 27 Feb 2015 1 Max Planck Institute for Mathematics in the Sciences, Inselstraße 22, 04103 Leipzig, Germany,
More information3 No-Wait Job Shops with Variable Processing Times
3 No-Wait Job Shops with Variable Processing Times In this chapter we assume that, on top of the classical no-wait job shop setting, we are given a set of processing times for each operation. We may select
More informationMathematical Programming and Research Methods (Part II)
Mathematical Programming and Research Methods (Part II) 4. Convexity and Optimization Massimiliano Pontil (based on previous lecture by Andreas Argyriou) 1 Today s Plan Convex sets and functions Types
More informationPS Geometric Modeling Homework Assignment Sheet I (Due 20-Oct-2017)
Homework Assignment Sheet I (Due 20-Oct-2017) Assignment 1 Let n N and A be a finite set of cardinality n = A. By definition, a permutation of A is a bijective function from A to A. Prove that there exist
More informationThe important function we will work with is the omega map ω, which we now describe.
20 MARGARET A. READDY 3. Lecture III: Hyperplane arrangements & zonotopes; Inequalities: a first look 3.1. Zonotopes. Recall that a zonotope Z can be described as the Minkowski sum of line segments: Z
More information